Custom Discovery Prompts
Your AI discovery prompt set is the set of questions used to measure whether AI assistants mention your app and its competitors. These prompts drive your AI discovery score, Share of Voice and intent coverage.
By default, each audit generates a calibrated set from your listing. Free and Basic plans use this automatic set without editing controls. Plans with prompt editing show an editable draft in the project’s AI configuration, alongside Brand Identity.
Customize your scored set
- Run an audit to generate suggestions. The editor seeds its draft from the latest audit’s generated prompt battery for the selected store and locale.
- Edit, add, remove or reorder prompts. Your plan’s prompt limit appears beside the editor; if the automatic battery is larger, the draft starts with the allowed number of prompts.
- Keep the prompts you want to measure active, then save. Saving replaces the automatic battery with your active prompts for subsequent audits and SoV tracking.
Saved prompts run exactly as written. A suggestion to phrase a prompt as a discovery question is advisory and does not prevent saving. The set stays fixed when you change your listing, until you edit it or reset it. You can also author a set before your first audit.
Reset to suggestions
Choose Reset to suggested set to clear the saved custom set. The next audit generates a fresh automatic battery. A previous custom audit is not treated as a new suggested set.
Add your own search terms
On plans with prompt editing, the project’s AI configuration also accepts your own search-demand terms — paste terms from a source like TikTok Studio, a TikTok Ads keyword planner, Google Search Console, or Apple Ads, one per line, and label where they came from. These terms are added to the search demand the engine harvests for your market and can ground new prompt suggestions, the same way a harvested phrase does. Removing a term you can no longer keep on a downgraded plan is always allowed, even past the plan’s limit.
Read the results
Results appear together in the main AI Discoverability view. Calibrated set identifies automatically generated prompts; Demand-grounded set identifies a prompt battery generated from harvested search demand; Your custom set identifies the prompts you saved. Use View prompt phrases to inspect the questions behind a result, including which search-demand phrase, if any, each prompt is grounded in. A suggested prompt you save keeps the demand match it was suggested with.
Share of Voice describes the tested prompts, not an objective cross-app benchmark. Score history and SoV tracking flag a change in how the prompt set was produced — for example switching between a custom set, a calibrated set, and a demand-grounded set — as not directly comparable. Ordinary automatic regeneration under the same method, including a fresh demand-grounded battery, does not produce that warning.